Keratinocyte stem cells (KSC) are believed to be identified by their ability to retain a BrdU label, and we have successfully FACS-sorted pure populations of these fixed and permeablelized label-retaining cells (LRC). In order to obtain living cells for biological studies, we have tried to identify unique cell surface markers on keratinocyte LRC, using both microarray analysis of global gene expression and proteomic mass spectrometry analysis of membrane proteins. For microarray analysis, we have developed navigated laser capture microdissection to isolate enriched populations of LRC in the human hair follicle and have identified panels of genes, including genes that encode membrane proteins, that are differentially expressed in the LRC population of the human hair follicle bulge area (CD200, Frizzled1 receptor, and follistatin) as compared to non-LRC keratinocytes (CD24, CD34, CD71, CD146). Using our in vivo competitive repopulation assay, we have recently characterized the stem cell behaviors of CD200+ keratinocytes present in non-hair bearing foreskins and in interfollicular epidermis. In collaboration with the Biomedical Proteomics Program at FCRF, we have developed high-throughput mass spectrometry methods able to identify membrane and cytoplasmic proteins on LRC, and we are now using semi-quantitative mass spectrometry to compare the relative levels of the KSC proteins prepared from LRC keratinocytes to proteins in control keratinocyte populations (transit amplifying basal keratinocytes) in order to identify unique panels of cell surface proteins on KSC and to detect novel or differentially expressed signaling pathways in KSC. We have also assessed stem cell behavior in other putative KSC populations including side population or SP cells that can be identified by unique fluorescent emission characteristics due to their ability to exclude HO33342 nuclear dye. We have recently described a SP population of keratinocytes and have completed characterization of SP keratinocytes long-term repopulating ability, using our recently described in vivo competitive repopulation stem cell assay. We have also now identified a sub-population of tumor initiating cells that are responsible for human squamous cell carcinoma. We have shown that this cellular subset is able to recapitulate the human skin cancer phenotype in immunocompromised mice and is highly enriched for tumor initiating cells. This subset is highly enriched for tumor initiating cells even though it only represents approximately 1% of total squamous cancer carcinoma cells. These tumor initiating cells can also be isolated from the tumors grown in immunocompromised mice and be used to serial transfer or passage the human squamous cell carcinoma to additional mice in a highly enriched fashion.